Editor’s Note: This post by Henry Aaron appears in tandem with a publication on the Health Affairs Web site of an
article by Jack Hadley of George Mason University, John Holahan of the Urban Institute, and coauthors,
which estimates the cost of covering uninsured Americans.
Back-of-the-envelope estimates of the added cost of covering the uninsured are easy to produce. Personal health care spending in 2008 will total an estimated $2 trillion. Approximately 17 percent of the population is uninsured. According to a commonly used rule of thumb, the uninsured consume about half as much health care as the insured.
But the uninsured on the average are younger and therefore use less care than do the insured. Seventeen percent, divided by two (to allow for care already consumed) and discounted for age differences brings the calculation down to about 6-7 percent of $2 trillion, or about $120-$140 billion.
A back-of-the-envelope estimate may be easy to make, but is it right? According to Hadley et al., the answer is ‘yes.’ Their estimate of $122.6 billion sits neatly at the lower end of that range. Although even $1 billion is ‘real money,’ as Senator Everett Dirksen famously observed, the $122.6 billion estimated cost of covering the uninsured is remarkably small—about the same as the growth of real health care spending over eighteen months, or about 0.85 percent of gross domestic product. How can something as important as universal coverage and so inexpensive compared to total health care spending remain so tantalizingly out of reach? The answer, which the authors of this article clearly understand, is even more important: to cover the uninsured, much or all health care financing must be restructured. (This answer, and the entire post, reflect my own views and not necessarily those of the Brookings Institution or any of its staff.)
Covering the currently uninsured has been a political ‘non-starter’ for decades and remains a long-odds proposition not because it is costly—it isn’t—but because it requires massive shifts in who writes the checks to pay for health care and who cashes those checks. Furthermore, the identity of the gainers and losers depends sensitively on how coverage is extended. The debate about extending coverage is not primarily about finding $120 billion to cover the uninsured, but about whether and how to shift who pays and who receives the many hundreds of billions dollars already being spent to continue covering the insured.
For a real-life display of what is at stake, one need go no further than the health programs of the two presidential candidates. Health reform proposals of Senators McCain and Obama would, if enacted, shift hundreds of billions of dollars in payments by businesses and households for premiums, cost sharing, and taxes and reallocate incomes of physicians, insurers, drug companies, and others. All the plans would differ from the status quo, but in very different ways.
The McCain Plan
Let’s start with Senator McCain. He proposes to repeal the current exclusion from income tax (but not the payroll tax) of employer-financed health insurance. Instead, taxpayers would be entitled to refundable tax credits—$2,500 for individuals and $5,000 for heads of household and joint filers. This is a genuinely radical proposal that would cause changes throughout the health care system and beyond, but in ways that no one can accurately anticipate. The income exclusion reduces federal revenues by an estimated $152 billion in fiscal year 2008; that cost is expected to reach $250 billion in five years. If all tax filing units were eligible for and applied for credits, the cost of the McCain plan would exceed $500 billion, more than three times the current revenue loss from the personal income tax exclusion of employer financed health insurance, but less than that amount plus outlays on Medicaid and Medicare. Computing the impact the refundable credits on public spending and on individual households and businesses is fraught with uncertainty. The following is a partial explanation of why gains and losses would abound but are hard to anticipate:
1) If the exclusion were removed, it is unclear whether employers would continue to sponsor and pay for health insurance. If they did so, it is not clear whether or how much of the increased net cost arising from the loss of deductibility they would shift to workers through lower wages, increased premiums, or restrictions on coverage. If they did not, it is unclear how much of the premium savings they would pass on to workers. The stakes and uncertainties on what happens to the hundreds of billions of dollars that employers now spend on health insurance are vast.
2) Presumably workers would receive the credit only if they were covered by insurance that met statutory standards. It is not clear what those standards would be.
3) If employers stop sponsoring health insurance, it is unclear whether workers would buy insurance individually or through newly-formed groups. In either case, premiums and insurance terms would vary geographically and by worker characteristics, including age and medical history. A nationally uniform credit would cover widely varying proportions of premium costs.
4) It is not clear how many workers now covered by employer-sponsored plans would elect to ‘go bare,’ simply pocketing any increase in wages that employers who dropped coverage might proffer and electing to gamble on staying healthy, being able to afford the costs of illness, or being able to fall back on whatever safety nets remained for the uninsured.
5) If employers dropped coverage and raised wages by the amount of reduced premiums, the increase in take-home pay would depend on each household’s marginal income tax rates. Personal income tax rates vary from –40 percent, for those in the phase up range of the earned income tax, to +35 percent, for those subject to the top personal income tax rate. (The maximum effective personal income tax rate is actually higher than 35 percent for those filers whose income puts them in the range where personal exemptions and itemized deductions are phased down.) Payroll tax rates vary from 2.9 percent to 15.3 percent depending on earnings levels, with the lower rate for workers who earn more than the Social Security wage base.
6) The McCain plan would likely induce insurers to offer new products and to reprice old ones, because individual purchasers would be exposed directly to the full incremental cost of insurance. These shifts would feed back to decisions of companies on whether to continue sponsoring and paying for insurance and to decisions of individuals on whether and what kind of coverage to buy.
7) Senator McCain has not made clear what changes, if any, would be made in Medicaid and Medicare or, more generally, what additional financial assistance might be made available to low-income households. The impact of the McCain plan on access to health care, the disposable incomes of the poor and near poor, and the finances of state and local governments would depend on policy decisions made by the federal, state, and local governments on how much help, in addition to the tax credits, would be provided to those now covered by Medicaid, SCHIP, and state-financed assistance to the poor and near poor.
Far more detail than has yet been released would be needed to model the impact of the McCain plan. Even then, estimates would depend sensitively on model parameters that cannot be reliably estimated because the plan would take the nation outside the range of past experience.
The Obama Plan**
Senator Obama proposes to move in a different direction. The income-tax exclusion would be retained. In addition, households would be offered a refundable tax credit equal to the excess of premiums on a specified plan over 7-½ percent of income. Employers who do not offer coverage or contribute to the cost of insurance for their employees that is ‘meaningful’ would be required to pay a 4 percent payroll tax to help subsidize coverage for low- and moderate-income families. SCHIP and Medicaid would be expanded. Insurance coverage would be mandated for children. The market for health insurance would be regulated through a national health insurance exchange. A variety of measures would be undertaken to reduce the growth of health care spending, some of which also appear in the McCain plan, including the promotion of competition and encouragement of information technology. Both candidates claim these and other measures would save a lot of money, but the Congressional Budget Office has been loathe to endorse such claims for similar initiatives when advanced in other contexts.
Who would gain and who would lose and how much if Senator Obama’s ideas were implemented is unclear. One must guess at important elements of the plan, such as how the mandate for children to be insured would be enforced. And, as with the McCain plan, estimates would hinge on assumptions about behavioral parameters estimated with data from an environment quite different from what would arise under the proposal.
What is clear is that the system-wide net cost of universal coverage does not begin to reveal the tidal shifts in spending that all plans necessary to achieve that goal entail. Nor does it hint at the redistribution among providers that would result from measures such as those proposed in different forms by both candidates to promote generic drugs or malpractice reform. On a day when scores of individual stocks hit new highs and lows, a nearly unchanged stock market index tells little of the underlying turmoil. In much the same way, the modest increase in system-wide costs needed to insure everyone reveals little about the political and economic obstacles that must be overcome to achieve it. Achieving universal coverage is mostly about income redistribution—among politically and economically powerful payers and providers with stakes that dwarf those measured by the added system-wide cost of insuring everyone.
** The specifics of the Obama plan are drawn from a memorandum prepared after conversations with the campaigns by Roberton Williams to guide modeling of the candidates’ plans by the Urban/Brookings Tax Policy Center. These specifics are not available at the campaigns’ web sites and are subject to change.